[uf-new] Currency brainstorming

[Guillaume Lebleu]

Andy Mabbett wrote:
> In message <46FD6147.5000603 at digitalbazaar.com>, Manu Sporny 
> <msporny at digitalbazaar.com> writes
>
>>> I
>>> believe the above example where the date of the currency is not 
>>> implicit
>>> is rare enough to be left aside for now for the purpose of moving this
>>> proposal forward and avoiding confusion within publishers.
>>
>> +1 - this seems like a fairly rare case (hardly any examples to date)
>
> Examples of such usage have been provided; past experience is that 
> quantitative evidence is of little interest to the community.
>
(Scott, we are indeed talking about 
http://microformats.org/wiki/currency-examples#Historic_prices, sorry 
the original link got discarded over the course of the thread).

Andy, I think the community has recognized that there are many cases 
where currency amounts used in the context of describing past events or 
past contents, items, etc.

What I disagree with is the idea that in the majority of these cases, 
the nearby date is a property of the money amount, as your suggestion of 
including a "date" property in "hmoney" suggests. I am instead arguing 
that in the majority of these cases, the nearby date is the property of 
another concept (not the "money amount" concept), be it the article, the 
price of the item described, or the price of a currency described. There 
aren't microformats or POSH practices for all of these concepts yet, but 
that's not a reason to push properties that belong to these 
yet-to-be-defined formats in the hmoney microformat.

In other words, in most of these cases, if needed, a parser can derive 
the "date" of the money amount by looking at the date of the article, 
item, etc. (provided a microformat is available for these).

In your example "In 1950 the average weekly wage was $X", I am arguing 
that "1950" is not a property of the currency amount X, but a property 
of the whole statement/paragraph. I believe we don't want publishers to 
mark up this content like <span class="hmoney">In <abbr class="date" 
title="...">1950</abbr> the ... was <abbr class="currency" 
title="USD"><span class="value">X</span></span>, just because we don't 
have an hStatement or hParagraph.

The cases where the date is indeed a property of the currency amounts is 
when we say "2005 dollars" or "1975 dollars". See 
http://www.google.com/search?q=%221975+dollars%22 (301 occurrences).

There are quite a few occurences of this last case indeed - try with 
other years (still few compared to the # of occurrences of currency 
amounts), but unless we can clearly reserve the use of date in hmoney 
for that special case, and not use it in the case of "In 1950 the 
average weekly wage was $X", then I'd rather leave the date property out 
of hmoney for now.

Have a nice week-end.

Guillaume

PS: Please refrain from using a tone such as in "past experience is that 
quantitative evidence is of little interest to the community". You don't 
attract flies with vinegar.